Copper and zinc fractions in the profile of an Inceptisol cultivated with apple in southern Brazil

Bragantia, Campinas, Ahead of print, 2018 ABSTRACT: The present study aimed to evaluate the accumulation and distribution of Cu and Zn fractions in the profile of an Inceptisol by application of fungicides with a previous history of apple cultivation in southern Brazil. Stratified soil samples (0.0 – 0.025 m, 0.025 –0.050 m, 0.05 – 0.10 m, 0.10 – 0.15 m, 0.15 – 0.20 m and 0.20 – 0.40 m depth) were collected in two apple orchards (8 years old and 18 years old), and in a native pasture area adjacent to the orchards. The soil was subjected to analysis of physical and chemical properties and to chemical fractionation of Cu and Zn. The fractions of soluble, exchangeable, organic, residual and total Cu and Zn accumulated Copper and zinc fractions in the profile of an Inceptisol cultivated with apple in southern Brazil Gustavo Brunetto1*, Lucas Benedet2, Vítor Gabriel Ambrosini2, Jucinei Jose Comin2, George Wellington Bastos de Melo3, Matheus Ademir dos Santos2, Cledimar Rogério Lourenzi2, Arcangelo Loss2, Paulo Belli Filho2, Djalma Eugênio Schmitt2, Rafael Couto1 1.Universidade Federal de Santa Maria Departamento de Solos Santa Maria (RS), Brazil. 2.Universidade Federal de Santa Catarina Centro de Ciências Rurais Florianópolis (SC), Brazil. 3.Empresa Brasileira de Pesquisa Agropecuária Uva e Vinho Bento Gonçalves (RS), Brazil.


ABSTRACT:
The present study aimed to evaluate the accumulation and distribution of Cu and Zn fractions in the profile of an Inceptisol by application of fungicides with a previous history of apple cultivation in southern Brazil.Stratified soil samples (0.0 -0.025 m, 0.025 -0.050 m, 0.05 -0.10 m, 0.10 -0.15 m, 0.15 -0.20 m and 0.20 -0.40 m depth) were collected in two apple orchards (8 years old and 18 years old), and in a native pasture area adjacent to the orchards.The soil was subjected to analysis of physical and chemical properties and to chemical fractionation of Cu and Zn.The fractions of soluble, exchangeable, organic, residual and total Cu and Zn accumulated

INTRODUCTION
In the state of Santa Catarina, in southern Brazil, an area of approximately 18.000 ha is cultivated with apple crops (Malus domestica).It is the largest apple-producing area in Brazil, and Santa Catarina is the second largest producer, after Rio Grande do Sul state.Current apple production amounts to 531.000 t, equivalent to 44% of national production (Epagri 2013; Acate 2014).Nevertheless, apple trees are grown in regions with frequent rainfall, which stimulates the incidence of foliar fungal diseases and may reduce fruit yield and harm fruit quality.Therefore, foliar fungicides, e.g., the copperbased Bordeaux mixture (Ca(OH) 2 + CuSO 4 ) and zincbased mixtures, such as those with the active ingredient Mancozeb (C 4 H 6 MnN 2 S 4 × Zn), are applied throughout the year.Furthermore, fungicides containing Cu and Zn are applied after leaf senescence and during plant dormancy to promote dormancy breaking and branches disinfection.
The accumulation of Cu and Zn fractions, especially in the upper soil layers, can be stimulated by several factors: unidirectional fungicide applications; Zn and Cu being washed off the leaf surface by rainfall; the deposition of senescent leaves and pruned branches; the cycling of Cu and Zn by species of ground cover plants that co-inhabit the orchards; a low transfer of Zn and Cu to apples; and an absence of soil disturbance between orchard rows (Fernández-Calviño et al. 2010;Mackie et al. 2012;Ruyters et al. 2013).
In soil, Cu and Zn are retained by physical and chemical bonds, by mineral and organic components, and their availability depends on: the mineral contents, oxides and hydroxides of Fe, Al and Mn; carbonates (Bradl 2004); soil pH values (Toselli et al. 2009); cation exchange capacity (CEC); content; and composition of organic matter (Fernández-Calviño et al. 2010).However, when Cu and Zn contents increase in the soil, as a result, for example, of frequent fungicide applications, there can be accumulation and changes in the distribution of their fractions in such soil components.Adsorption of Cu and Zn in the soil occurs primarily in high affinity sites, usually forming inner sphere complexes.The remainder is adsorbed at sites with low binding energy and, consequently, it increases the availability and mobility of these elements in the soil profile (Brunetto et al. 2014).Thus, the increase of Cu and Zn in the soil over the years may increase the amount of soluble and exchangeable fractions of these elements, increasing toxicity to plants (Michaud et al. 2007;Nagajyoti et al. 2010;Yadav 2010;Fernández-Calviño et al. 2012) and water contamination through their transfer in the soil profile (Mileusnić et al. 2014;Marchand et al. 2014).
The fractions of Cu and Zn can be estimated by chemical fractionation, which removes Cu and Zn from the most labile to the most stable fractions (Tessier et al. 1979).The total amount of these elements in soils or sediments can be separated into bioavailable fractions (water-soluble and exchangeable), potentially bioavailable fractions (bound to clay minerals, oxides, carbonates and organic matter) and residual fractions (bound to the structure of minerals) (Casali et al. 2008;Frankowski et al. 2010;Brunetto et al. 2013;Duplay et al. 2014).However, in the Americas, particularly in subtropical and temperate regions, few studies exist on the accumulation and distribution of Cu and Zn fractions in the soil profile in apple orchards with a history of continuous fungicides application, whose compositions include Cu and Zn.The present study aimed to evaluate the accumulation and distribution of Cu and Zn fractions in the profile of an Inceptisoil on two plantations previously cultivated with apple trees in southern Brazil.

Study location
This study was conducted using soil collected from two commercial orchards, one with eight years of cultivation (O8) and another with 18 years of cultivation (O18), and a soil from native pasture (NP), located nearby the orchards, in the county of Urubici, in the Planalto Serrano region, state of Santa Catarina, southern Brazil (Long 49°35'30"W, Lat 28°00'05"S).At the three sites, the soil was classified as an Inceptisoil, a Cambissolo Húmico in Brazilian classification (Embrapa 2013).The local relief was classified as slightly undulating, with a 4% slope.The climate is humid mesothermal with mild summers, Cfb type according to Köppen classification, with an average annual rainfall of 1360 -1600 mm, average temperature of 15.4 °C, and between 20 and 29 frosts can occur throughout the year (Pandolfo et al. 2002).
Orchard O8 was implanted in 2003 with a planting density of 4,444 plants•ha -1 (4.5 m between rows and 0.5 m between plants), and orchard O18 in 1993, with a planting density of 1,482 plants•ha -1 (4.5 m between rows and 1.5 m between plants).The orchards had two commercial varieties, with Gala and Fuji accounting for 70% and 30% of the plants, respectively, with trees grown under a central-leader planting system.The plants of orchard O8 were grafted onto a Marubakaido rootstock, with a 20-cm M9 filter, while in orchard O18 the plants were grafted onto a Marubakaido rootstock without filter.
The implantation of the orchards was preceded by an application of lime to raise the soil pH in water to 6.0.Lime was applied on the surface and incorporated into the soil with a plow and then with a harrow to a depth of approximately 0.3 m.The plantation received 1000 kg•ha -1 of formula fertilizer (5-20-10; N-P 2 O 5 -K 2 O), totaling 50, 200 and 100 kg•ha -1 of N, P 2 O 5 and K 2 O, respectively.The fertilizer was applied to the soil surface and also incorporated to approximately 0.3 m.Maintenance fertilizer was added annually: approximately 50, 30 and 150 kg•ha -1 of N, P 2 O 5 and K 2 O in the formulation 11-7-35.The fertilizer was applied on the soil surface without incorporation, in a row coincident with the canopy of the plants.Weed control in these rows was performed annually with the use of an herbicide (with glyphosate as the active ingredient).Between rows, weeds consisted primarily of Trifolium repens, Trifolium pratenses, Paspalum notatum and Chaptalia nutans.They were mowed throughout the apple tree cycle at approximately 0.10 m height from the soil surface.The native pasture area, which was used as reference, was located adjacent to the orchards, and natural vegetation was composed mainly of cover crops.
For fungal disease control during the production period of the orchard (October to April), was applied a spray volume of approximately 1000 L•ha -1 of fungicides N-(4,6-dimethylpyrimidine-2-yl) aniline (PYRIMETHANIL) (45 mL 100 L); N-(trichloromethylthio) cyclohex-4-ene-1,2-dicarboximide (CAPTAN) (22 g 100 L); Sulfur (sulfur) (100 g 100 L); and especially Mancozeb (160 g 100 L).The fungicides were applied every seven days, always with rotation between the products.However, if it rained for more than nine consecutive hours, a new application was performed.During the winter period, the plant apple trees received 1000 L•ha -1 spray volume of a copper-based fungicide (Copper II + Oxychloride -Cu oxychloride), one application at the beginning of the production cycle, during the budding of the leaves, and other at the end of the cycle, after leaf senescence, with a total of around 22 fungicide applications per year.

Collection and preparation of soil samples
In October 2011, eight years after the implantation of orchard O8 and 18 years after the implantation of orchard O18, six trenches measuring 0.5 m depth and 0.5 × 0.3 m side were opened in random spots close to the planting rows in each orchard.Six trenches were opened at random points on the NP.Stratified samples were collected in each of the trenches in the following layers: 0.0 -0.025 m, 0.025 -0.05 m, 0.05 -0.10 m, 0.10 -0.15 m, 0.15 -0.20 m and 0.20 -0.40 m.Soil samples were air-dried, ground in a porcelain mortar, passed through a 2-mm mesh sieve, and reserved for analysis.

Physical and chemical characterization of soil
The soil samples were quantified for the levels of clay, silt and sand by the pipette method (Embrapa 1997).Total organic carbon was determined by the Walkley-Black method (Embrapa 1997), and the levels of soil organic matter were measured by multiplying the total organic carbon content by 1,724 (Tedesco et al. 1995).Analysis of soil pH was performed using a soil:water suspension of 1:1 v/v ratio (Tedesco et al. 1995).The levels of available phosphorus (P) and exchangeable potassium (K) were determined in a solution extracted by Mehlich -1 solution (HCI 0.05 mol•L -1 + H 2 SO 4 0.0125 mol•L -1 ); P was determined by colorimetry (Murphy and Riley 1962) in a spectrophotometer; and K in a flame photometer.The contents of exchangeable Ca 2+ , Mg 2+ and Al 3+ were determined in a solution extracted by KCl (1 mol•L -1 ), whereas Ca and Mg were determined by atomic absorption spectrometry (AAS), and Al by titration with NaOH (Tedesco et al. 1995).Effective cation exchange capacity (CEC effective ) was calculated by summing the exchangeable cations of Al 3+ + Ca 2+ + Mg 2+ + K + .Cation exchange capacity at pH 7.0 (CEC pH 7.0 ) was calculated by the sum of exchangeable bases (Ca 2+ + Mg 2+ + K + ) plus H + Al SMP estimated from the equation proposed by Kaminski et al. (2002).Table 1 shows the determined physicochemical parameters.

Chemical fractionation of copper and zinc in soils
The chemical fractionation of Cu and Zn was carried out by adapting the methodology proposed by Tessier G. Brunetto et al. et al. (1979).Soil samples (1 g, in triplicate) were transferred to 50 mL polypropylene tubes to perform the following sequential extractions: the soluble fraction, extracted with deionized water (8 mL) (Cu sol and Zn sol ); the exchangeable fraction, extracted with 1.0 mol.L -1 MgCl 2 at pH 7.0 (8 mL) (Cu E and Zn E ); the mineral-bound fraction, extracted with 0.04 mol.L -1 hydroxylamine hydrochloride (NH 2 OH.HCl) in 25% acetic acid (v/v) at pH 2.0 (20 mL) (Cu Min and Zn Min ); and the organic-bound fraction extracted with 0.02 mol.L -1 HNO 3 (3 mL) + 30% H 2 O 2 (8 mL) + NH 4 OAc in 20% HNO 3 (v/v) (5 mL) (Cu OM and Zn OM ).The supernatant was separated by centrifugation at 3200 xg for 30 min, and an aliquot was set aside to determine the Cu and Zn levels.Between each extraction, the samples were agitated with deionized water and centrifuged, and the supernatant was discarded.At the end of the sequential extraction, the remaining residues were dried at 65 °C for 24 h, and then, 0.50 g of the dry residue was subjected to digestion with concentrated HF + HClO 4 to determine the residual Cu and Zn fractions (Cu R and Zn R ).To determine the pseudototal levels, 30% H 2 O 2 + concentrated HF + concentrated HClO 4 solution was added to 0.5 g samples of dry soil (not subjected to sequential extraction of Cu and Zn) to digest the samples (Cu T and Zn T ).In conducting the study, all glassware was soaked in 10% HNO 3 for at least 24 h and then washed in deionized water three successive times before use.The Cu Sum and Zn Sum represent the sum of fractions Cu Sol , Cu E , Cu Min , Cu OM , and Cu R and Zn Sol , Zn E , Zn Min , Zn OM , and Zn R, respectively.

Statistical analysis
The results of the Cu and Zn fractions were submitted to analysis of variance (ANOVA), and when the effects were significant between the layers of the same area, the means were compared by Tukey's test (a = 5%).The differences between areas were evaluated using orthogonal contrasts: native pasture versus orchards and orchard O8 versus orchard O18 within each sampled soil layer.An analysis of variance was performed, and the differences in the proposed contrasts were tested at 5%, 1% and 0.1% probability of error for the F-test.
The fractions of Cu Sol , Cu E , Cu Min , Cu OM , Cu R and Cu Sum and the fractions of Zn Sol , Zn E , Zn Min , Zn OM , Zn R and Zn Sum were submitted to principal component analysis (PCA).For this purpose, the mean of six replicates was calculated for each layer and subjected to PCA.To minimize the discrepancy between the data, the values were normalized using Eq.1: (value -minimum value / maximum valueminimum value) × 2 -1 PCA is used to reduce data size and thus facilitate analysis by means of a correlation circle graph (Herlihy and McCarthy 2006).

Fractions of Cu in soil
In the native pasture (NP) soil, the highest levels of Cu Sol , Cu R , Cu Sum and Cu T were observed in the layer of 0.0 -0.025 m and decreased at deeper levels (Table 2).The levels of Cu E in the NP soil did not differ between layers.The highest levels of Cu Min in the NP soil were found in the layers at 0.0 -0.025 m and 0.025 -0.050 m, whereas for the Cu OM fraction, the higher content was observed in the layers at 0.0 -0.025 m to 0.10 -0.15 m.In the soil of O8 and O18 orchards, the highest levels of all Cu fractions were observed in the layer at 0.0 -0.025 m and decreased in deeper layers.
The Cu R fraction had higher levels compared with the Cu Sol, Cu E, Cu Min , and Cu OM fractions in all soil layers of NP (Table 2).The levels of Cu Sol , Cu E , Cu Min and Cu OM were similar to each other, except for the layer at 0.15 -0.20 m, in which the Cu E fraction had higher levels than the fractions of Cu Sol , Cu Min and Cu OM (Table 2).In all soil layers evaluated for O8, the levels of Cu R were also higher than those found in the Cu Sol , Cu E , Cu Min and Cu OM fractions, with the exception of the layer at 0.0 -0.025 m, where there were no differences in the level of Cu OM .For the orchard O18, there were also higher levels of Cu E compared with the Cu Sol , Cu E , Cu Min and (1) Cu fractions (mg Cu OM fractions only in the layers at 0.05 -0.10 m and 0.10 -0.15 m.In the soil layer from 0.0 -0.025 m in orchard O18, the level of Cu Org was higher than the levels of Cu R , Cu Min , Cu E and Cu Sol (Table 2).
The means of the levels of all Cu fractions in the six evaluated layers (0.0 -0.025 m, 0.025 -0.05 m, 0.05 -0.10 m, 0.10 -0.15 m, 0.15 -0.20 m and 0.20 -0.40 m), except for the Cu E fraction, were higher in the soil of orchards O8 and O18, compared with the levels of soil under natural pasture (Tables 2 and 3).This result was also observed when comparing the means of Cu levels in the six evaluated layers between the two orchards, where all fractions, except for Cu E , were higher in the orchard O18 compared with O8 (Tables 2 and  3).Furthermore, the Cu Min , Cu OM , Cu E and Cu sum fractions were higher in the orchard O18 compared with O8 in the six layers (Tables 2 and 3).The exceptions were the levels of fractions Cu E and Cu Total (0.025 -0.05 m), Cu Sol (0.05 -0.10 m) and Cu Sol and Cu Total (0.10 -0.15 m and 0.15 -0.20 m), which showed no differences (Table 3).
The native pasture layer at 0.0 -0.025 m had the highest percentage of Cu in the residual fraction (Cu Res ) (Figure 1), but in the soil of O8 and O18, the percentage of Cu Res decreased compared with that of the native pasture soil.There was also an increase in the Cu OM and Cu Min fractions.In the soil of O18, the percentage of Cu OM was higher than that of the Cu E and Cu Res fractions.In the layer at 0.025 -0.050 m, the Cu Res percentages decreased in the soil of O8 and O18, compared with that of the native pasture soil, because of an increase in the percentage of Cu OM , especially in the soil of O18.In the layer at 0.05 -0.10 m, the percentages of Cu Res were similar between the soil under native pasture and that of O8.However, in the soil of O8, the percentages of Cu OM and Cu Min were higher compared to the soil under native pasture.The highest percentages of Cu OM and Cu Min were observed in the soil of O18, compared with the soils of the native pasture and O8.In the layers at 0.10 -0.15 m, 0.15 -0.20 m and 0.20 -0.40 m, the percentages of Cu OM and Cu Min increased in the soil of O8, and especially in the soil of O18.In the latter, because of the increase in the Cu OM and Cu E fractions, a lower proportion of Cu Res was present compared with NP and O8.For the Cu E fraction, the soil under native pasture had the highest percentages in all evaluated layers.
The principal component analysis (PCA) of the Cu fractions showed a relation between the main component 1 (factor 1) and the main component 2 (factor 2), which was split among the areas (NP, O8 and O18) (Figure 2a), accounting for 98.87% of the results.The areas of native pasture and O8 were separated from O18 by factor 1 of PCA, which helps to explain 95.4% of the results.The result was strongly influenced by the variables Cu sol , Cu E , Cu Min , Cu OM and Cu Res .Factor 2 of the PCA, which helps explain only 3.4% of the results, separated the areas of orchards O8 and O18 from the native pasture area.

Zn fractions in soil
The highest levels of Zn fractions (Zn Sol , Zn E , Zn Min , Zn OM , Zn R and Zn Sum ) and Zn Total were found in the layer at 0.0 -0.025 m in all areas assessed, with a subsequent decrease in the levels because of increasing depth (Table 4).The exception to this pattern was found in the soil of O8 and O18, where the Zn OM and Zn Total fractions did not differ between the layers at 0.0 -0.025 m and 0.025 -0.050 m.Furthermore, in the soil of O18, the Zn Res fraction showed equal values between those two layers and the layer at 0.0 -0.15 m. 1 Means of six evaluated layers.*Significant at 5%; ** significant at 1%; *** significant a 0.1%; ns = no significant.In all layers and areas assessed, the highest Zn levels were found in Zn Total , except for the Zn sum fraction, which showed values equal to those of the Zn Total fraction in the layer at 0.05 -0.10 m for O8.For the Zn Res fraction, there were higher levels compared with the Zn Sol , Zn E , Zn Min and Zn OM fractions in all layers of the soil under natural pasture, O8 and O18.The level of the Zn Min fraction in the soil under native pasture, O8 and O18 in all layers was lower than that of Zn Res , but higher than the levels of Zn Sol , Zn E and Zn OM .In the soil of O18, in all layers, the Zn E levels were higher than the levels in the soil layers of the native pasture and O8.
Contrast orthogonal analysis showed that the mean Zn content of the six measured layers (0.0 -0.025 m, 0.025 -0.05 m, 0.05 -0.10 m, 0.10 -0.15 m, 0.15 -0.20 m and 0.20 -0.40 m) and of all fractions (Zn E , Zn Min , Zn OM , Zn R , Zn Sum and Zn Total ) was higher in the soil of O8 and O18, compared with the soil under native pasture (Tables 4 and 5).This result was also found by comparing the means of Cu levels measured in the six layers between the two orchards, where all the fractions were higher in orchard O18 compared with O8.Except for the Zn Sol fraction in the layers at 0.10 -0.15 m and 0.20 -0.40 m, and Zn R in the layer at 0.0 -0.025 m, all other fractions in all layers were higher in the soil of O18, compared with that of O8 (Tables 4 and 5).

Layer (m)
Zn fractions (mg Means of six evaluated layers.*Significant at 5%; ** significant at 1%; *** significant a 0.1%; ns = no significant. In the layers at 0.0 -0.025 m, 0.025 -0.050 m, 0.05 -0.10 m, 0.10 -0.15 m, 0.15 -0.20 m and 0.20 -0.40 m, the highest percentages of Zn in soil under the native pasture, O8 and O18 were found in the residual fraction (Zn Res ), being that the highest percentages were in the soil of O8 (Figure 2).However, in the soil of O18 was an increase of Zn content in the mineral fraction (Zn Min ) in all layers, compared with the soils of the native pasture and O8.
The PCA of Zn fractions showed the relation between the main component 1 (factor 1) and main component 2 (factor 2), which suggested separation of the areas assessed (Figure 2b), supporting the PCA of Cu fractions.Factors 1 and 2 accounted for 98.96% of the results.Factor 1, which explained 93.19% of the data, separated the areas of native pasture and O8 from that of O18.The Zn Sol , Zn E , Zn Min , Zn OM and Zn Res fractions contributed greatly to this result, whereas factor 2 contributed only 5.7% to the result and separated O8 and O18 from the native pasture area.

DISCUSSION
The highest levels of Cu fractions (Cu Sol , Cu E , Cu Min , Cu OM , Cu R , Cu Sum and Cu Total ) were observed in the surface layer (0.0 -0.025 m), especially in the soils of orchards O8 and O18.They can be attributed to years of applications of foliar fungicides, such as Mancozeb, Pyrimethanil and [CuSO 4 •5H 2 O + Ca(OH) 2 )], with the aim of preventing fungal diseases and also to promote dormancy breaking of plants (Verdisson et al. 2001;Komárek et al. 2010;Duplay et al. 2014).Thus, Cu may have accumulated on the soil surface because the applications were directional to the trees and Cu was washed off the surface of leaves by rainfall, and also by deposition of senescent leaves and pruned branches containing Cu derived from these applications (Brunetto et al. 2013).Furthermore, the higher level of Cu on the soil surface was also favored by the absence of soil disturbance in planting rows and between rows of apple trees over the years, by the cycling of Cu through deposition and subsequent decomposition of senescent leaves and pruned branches of apple trees, or even by the decomposition of roots in the soil (Amery et al. 2007;Yuan et al. 2007;Wang and Mulligan 2013;Pérez-Esteban et al. 2013).Therefore, the organic matter content in the upper layers of the soil are maintained and some Cu remains on the soil surface because its electron configuration is [Ar]3d 10 4s 1 , thus making it highly reactive with the functional groups of organic matter that contain S and N, in addition to the carboxyl and phenolic functional groups (Croué et al. 2003).Furthermore, organic compounds provide Cu with high binding energy and low desorption (Lair et al. 2006;Casali et al. 2008;Brunetto et al. 2013).
In the case of the orchards, especially O18, the higher Cu content, represented by Cu Total , can be attributed to the 22 annual fungicide applications throughout the years.In addition to the factors mentioned above, the fresh pulp of apple fruit is between 2 and 4 mg•kg -1 Cu (Stüpp et al. 2013).In this sense, an orchard with production of 60 t•ha -1 has exported 0.24 kg•ha -1 Cu, a fact that contributes to a higher accumulation of this trace element in the soil.Cu accumulation in the orchards was also favored because the relief at the site of the orchards is gently undulating, and cover crops are planted between the orchard rows, favoring the deposition of plant residues on the soil surface, which minimizes the transfer of Cu by the drained solution (Loures et al. 2006;Pinho et al. 2006).The higher Cu content in the soil -mostly in the organic fraction (Cu OM ), compared with other fractions (Cu Res , Cu Min , Cu E and Cu Sol ), especially in the surface layer (0.0 -0.25 m), with increasing age of the orchards (eight years for O8 and 18 years for O18) and, therefore, a larger number of Cu-containing fungicide applications and higher nutrient-cycling time -can be attributed to high affinity and reactivity with the functional groups of the soil organic matter (Pérez-Novo et al. 2008).Therefore, because Cu is part of the organic fraction, the adoption of conservation management practices in orchards is crucial, for example, the deposition and maintenance of plant residues on the soil surface helps maintain the levels of soil organic matter and the absence of soil disturbance minimizes the mineralization of organic matter (Ferreira et al. 2013;López-Piñeiro et al. 2013).Nevertheless, the migration of the Cu to a depth of 0.40 m in the soil of orchards O8 and O18 was observed through the level of Cu Total and Cu OM , Cu Min and Cu Res fractions, (Table 3, Figure 1).This may have happened due to the saturation of functional groups of organic matter (Besnard et al. 2001;Croué et al. 2003), by the saturation of functional groups of Fe and Mn oxides (Bradl 2004), and the silanol and aluminol groups of phyllosilicates (Sposito 1989;McBride 1994).Furthermore, the decrease in pH values in the soil profile (Table 1) may change the sorption capacity of the clay minerals, especially iron oxides (Sposito 1989).At acidic pH values, sorption is lower and desorption is facilitated (Brunetto et al. 2013).
The highest levels of Zn Total , in particular, as has occurred with Cu in the most superficial soil layer (0.0 -0.025 m), mainly in the soil of orchards O8 and O18, are due to the unidirectional annual applications of fungicides containing Zn in their composition, such as Mancozeb (Komárek et al. 2010), Zn being washed off the surface of leaves by rainfall, and deposition of Zn-containing leaves or pruned branches (Fernández-Calviño et al. 2012;Brunetto et al. 2014;Duplay et al. 2014).Furthermore, the higher accumulation of Zn Total in the surface layers of the soil of the orchards, particularly orchard O18, is associated with a larger number of Zn-containing foliar fungicide applications for the preventive control of fungal diseases over the years (Brunetto et al. 2013).As seen for Cu, accumulation of Zn forms on the surface is also associated with an absence of soil disturbance in the rows and between the rows of apple trees, in addition to Zn cycling in the soil by the ground cover plants located in the rows and between rows of orchards (Couto et al. 2015;Hristovski et al. 2014).
In all soil layers of NP and orchards O8 and O18, higher levels of Zn were observed in the Zn Res and Zn Min fractions (Table 4 and Figure 3b), and these fractions are more stable and characterized by low availability and mobility in soil (Ramos 2006;Fernández-Calviño et al. 2012).Zn Total and the Zn E , Zn Min and Zn R fractions (Table 4 and Figure 3b) showed migration of Zn to 0.40 m depth.This may have occurred because of the migration of Zn by the flow of water in the soil because a fraction of Zn tends to remain in the soil solution in free form or as soluble ionic pairs (Citeau et al. 2003) and also by turbulent water flow in macropores and by soil cracks or even bioturbation (Girotto et al. 2010;Brunetto et al. 2013).In addition, Zn can bind to organic acids with low molecular weight, increasing its mobility in the soil profile (Wang and Mulligan 2013).This may occur because the formation of metal complexes with organic compounds with low molecular weight alters the balance between the metals in the labile form in the solid phase and in the soil solution, thus, dissolved organic compounds can also increase the release of Cu in the solid phase, and consequently increase its mobility in the soil (Girotto et al. 2010).In addition, Zn mobility may have been facilitated because of the decrease in water pH in the deeper layers of the soil, contributing to the formation of stable (innersphere) complexes between Zn and the functional groups of organic and mineral fractions of soil (Brunetto et al. 2013).
The increase in the Zn E fraction in the soil profile, which is a bioavailable fraction (Tessier et al. 1979) especially in the orchard O18 (Table 4), increases the potential for contamination of subsoil water (Banas et al. 2010;Fernández-Calviño et al. 2012).However, the increase in Zn E in the upper layers, where the roots of apple trees and ground cover plant species are located in the rows and inter-rows of the orchards, may potentiate toxicity to plants (Nagajyoti et al. 2010;Fernández-Calviño et al. 2012).One example is the data obtained by Zhang et al. (2013) and Hansi et al. (2014).The first authors, studying the length of barley (Hordeum vulgare) roots grown in soil in a greenhouse with the addition of 0, 12.5, 25, 50, 100, 200, 400 and 800 mg Cu•kg -1 , observed that root elongation drastically decreases with increasing Cu in solution.
Decreases of roots systems is not desirable because it reduces the nutrient cycling by ground cover plant species and the amount of plant residues deposited on the surface of soil, which are responsible for breaking the kinetic energy of raindrops, as well as reducing transportation of the solution on the soil surface.Moreover, it helps maintain the organic matter content in the soil (Brunetto et al. 2013).The increased Zn concentration in the soil can also cause toxicity to microorganisms, adversely affecting, for example, the decomposition of the crop residue deposited on the ground and the mineralization of soil organic matter (Gillera et al. 1998).Zn E also increases in the most superficial soil layer, which can enhance the transfer of Zn on the soil surface by the solution drained or adsorbed in the sediment (Fernández-Calviño et al. 2008;Fernández-Calviño et al. 2012), which, in turn, enhances contamination of surface water adjacent to the orchards.

CONCLUSION
The Cu and Zn fractions increased in the surface soil layer (0.0 -0.025 m) of apple orchards compared to native pasture cultivation, being associated with the time of cultivation.
Most Cu was distributed in the less stable organic fraction in the upper layers of apple orchards, which may potentiate toxicity to plants and transfer Cu to subsurface waters.In deeper layers, however, Cu was also distributed in the residual and mineral fractions, which are more stable.Most Zn in the soil of apple orchards was distributed in the residual and mineral fractions, indicating low mobility, as well as potential toxicity to plants.

Figure 1 .
Figure 1.Relative percentage of each Cu fraction obtained from a sequential extraction compared with the sum of the amount extracted from six layers of an Inceptisoil under native pasture and in orchards cultivated with apple trees (O8 and O18).

Figure 2 .Factor
Figure 2.The ratio between the principal component 1 (factor 1) and principal component 2 (factor 2), mean levels of (a) Cu and (b) Zn in the layer at 0.0 -0.40 m.Cu sol and Zn Sol -soluble fraction extracted with H 2 O; Cu E and Zn E -exchangeable fraction extracted with MgCl 2 1.0 mol•L -1 ; Cu Min and Zn Min -mineral fraction extracted with NH 2 OHHCl 0.04 mol L -1 in 25% CH 3 COOH; Cu OM and Zn OM -organic fraction extracted with HNO 3 0.02 mol•L -1 + 8 mL of 30% H 2 O 2 ; Cu Sum and Zn Sum -sum of all fractions; and Cu Res and Cu Res -residual fraction extracted with HNO 3 + HCL + H 2 O 2 .

Figure 3 .
Figure 3. Relative percentage of each Zn fraction obtained from a sequential extraction compared with the sum of the amount extracted from six layers of an Inceptisoil under native pasture and in orchards cultivated with apple trees (O8 and O18).

Table 1 .
Physical and chemical attributes of an Inceptisoil under native pasture (NP) and in orchard O8 (8 years old) and orchard O18 (18 years old), both cultivated with apple trees.

Table 2 .
Cu fractions of an Inceptisoil under native pasture (NP) and in orchard O8 (8 years old) and orchard O18 (18 years old), both cultivated with apple trees.
1Means followed by the same capital letter in a column and lower case letter in a row do not differ by Tukey's test (p < 0.05).CV = coefficient of variation.

Table 3 .
Orthogonal contrasts between the Cu fractions of an Inceptisoil under native pasture (NP) and in orchard O8 (8 years old) and orchard O18 (18 years old), both cultivated with apple trees.

Table 4 .
Zn fractions of an Inceptisoil under native pasture (NP) and in orchard O8 (8 years old) and orchard O18 (18 years old), both cultivated with apple trees.

Zn Sol Zn E Zn Min Zn OM Zn Res Zn Sum Zn TotalTable 5 .
Orthogonal contrasts between Zn fractions of an Inceptisoil under native pasture (NP) and in orchard O8 (8 years old) and orchard O18 (18 years old), both cultivated with apple trees.